The present review considers the background to terminology that relates foods, glycaemia and health, including ‘available carbohydrate’, ‘glycaemic index’ (GI), 'glycaemic glucose equivalent', 'glycaemic response index' and 'net carbohydrate', and concludes that central to each of these terms is 'glycaemic load' (GL). GL represents the acute increase in exposure of tissue to glucose determined by foods; it is expressed in ingested glucose equivalents (per 100 g fresh weight or per serving), and is regarded as independent of the state of glucose metabolism from normal to type 2 diabetes mellitus (T2DM). Ad libitum studies in overweight or obese adults and children show that low-GL diets are associated with marked weight benefits, loss of adiposity and reduced food intake. Weight benefits appear on low-glycaemic v. high-glycaemic available carbohydrates, unavailable v. available carbohydrates and protein v. available carbohydrate. Energy intake immediately after lowering of meal GL via carbohydrate exchanges is apparent only after a threshold cumulative intake of >2000 MJ. Various epidemiological and interventional studies are discussed. A relationship between GL and the development of T2DM and CHD is evident. Studies that at first seem conflicting are actually consistent when data are overlaid, such that diets with a GL of >120 glucose equivalents/d would appear to be inadvisable. Whereas certain studies might place GI as being slightly stronger than GL in relationto T2DM risk, this situation appears to be associated with observations in a lower range of GL or when the range of GI is too narrow for accuracy; nevertheless, authors emphasise the importance of GL. Among the studies reviewed, GL offers a better or stronger explanation than GI in various observations including body weight, T2DM in nurses, CHD, plasma triacylglycerols, HDL-cholesterol, high-sensitivity C-reactive protein and protein glycation. Where information is available, the associations between risk factors and GL are either similar or stronger in the overweight or obese, as judged by BMI, and apply to both body weight and blood risk factors. The implications tend to favour a long-term benefit of reducing GL, for which further study is necessary to eliminate any possibility of publication bias and to establish results in clinical trials with overweight and obese patients.